Dongseok Kwon

Title: Innovations of Dongseok Kwon in Neuromorphic Architecture

Introduction

Dongseok Kwon is a prominent inventor based in Seoul, South Korea. He has made significant contributions to the field of neuromorphic systems, particularly through his innovative patent that focuses on on-chip training architectures. His work is pivotal in advancing the capabilities of neural networks.

Latest Patents

Dongseok Kwon holds a patent for an "On-chip training neuromorphic architecture." This neuromorphic system enables on-chip training and includes several key components. It features synapse arrays arranged in a cross-bar shape, a final neuron layer with both forward and backward neurons connected to an output terminal, and neuron layers that store signals used during weighted value updates. The system also incorporates an error calculation circuit that detects and outputs error values, enhancing the performance of neural networks. The conductances of the synapse devices represent the weighted values and are modified through a weighted value update operation. Each synapse device is configured with a flash device, and the neuron layers are implemented with ultra-miniature devices.

Career Highlights

Dongseok Kwon is affiliated with Seoul National University, where he continues to push the boundaries of research in neuromorphic systems. His work has garnered attention for its innovative approach to integrating hardware and neural network training.

Collaborations

Dongseok Kwon collaborates with notable colleagues, including Jong-Ho Lee and Jangsaeng Kim. Their combined expertise contributes to the advancement of research in their field.

Conclusion

Dongseok Kwon's contributions to neuromorphic architecture represent a significant step forward in the integration of artificial intelligence and hardware. His innovative patent showcases the potential for on-chip training systems to revolutionize neural network applications.